Generally it can't if it is cellular. Air may be replaced with a different gas though and that may be less conductive than air.
Vacuum gets around it.

If air movement happens then your into a different calculation, why shape is also important.

All these insulants are basically ways of immobilising air and preventing convection.

Perhaps the U value for air refers to free air and assumes some convection, though I have no idea how you can do that and obtain a consistent U value. U value calculations are appropriate for conduction only, and if you bring in convection (free air) or radiation (multifoil) you won't get consistent results.

Posted By: fostertomAerogel, better than 0.0241 but full of nothing but air?

Has to do with getting the cell size smaller than the distance that air molecules travel before bashing into another one. Sadly, I don't remember the precise reason, only that there is a real one.

graphited EPS?

Doesn't have a conductivity less than 0.024 AFAIK.

Posted By: djhSadly, I don't remember the precise reason,

I think the idea is that fast-moving (hot) molecules bounce off a cell wall before before they can carry their energy very far.

At atmospheric pressure the mean free path of air molecules is only 68nm (Wikipedia), a fraction of the wavelength of light. Only in materials like aerogel, which have an invisibly fine-grained porous structure, will many molecules hit the solid matrix before they hit another gas molecule.

In coarser materials most collisions will be with other gas molecules. But momentum will be transferred, and the next molecule will be knocked in the same general direction. So it still helps to put something in the way which is sufficiently rigid and massive to bounce 'hot' molecules back the way they came. That, I think, is why coarser, foam-type materials can still achieve conductivities below that of free air.

All the above refers only to conduction, i.e. heat transfer by random molecular motion. Insulants additionally suppress convection, where heat is transferred by bulk movement of air.

Posted By: rhamdurefers only to conduction, i.e. heat transfer by random molecular motion. Insulants additionally suppress convection

What about insulants that suppress internal radiation (multifoils), or ones that encourage the radiant portion of the heat flow to get absorbed into solids (graphited) and/or 'lost' in maze-like solids (aerogel)?

Posted By: SaintWho says air is the best insulant?

Not me, guvnor. The higher the molecular mass of the gas, the less heat it transfers. So argon is better than air in double glazing, and krypton is better still.

The blowing agents in phenolic and PIR typically start with four or five carbon atoms (already twice the molecular mass of O2 or N2) and weigh the molecule down with as much fluorine and chlorine as the ozone layer will tolerate. Good point, Saint, this answers how these materials outperform air.

Posted By: rhamduThe higher the molecular mass of the gas, the less heat it transfers. So argon is better than air in double glazing, and krypton is better still.

Isn't that about stopping convection, though?

Well yes, vs air at 24x10⁻³ W·m⁻¹·K⁻¹ so quite a bit better. My understanding, though, was that the heavier gasses were less likely to convect so you could have a thicker gap. However, looking around the descriptions I can find are pretty muddled. Still, it seems like I had the wrong end of the stick - the better resistance to conduction of the heavy monatomic gasses allows, for a given amount of conduction, a thinner gap hence less problem with convection.

U-value depends on thickness (OK, it's late).

Link to BBA certificate at bottom of this page:

http://www.icynenesprayfoam.co.uk/icynene-classic-spray-foam-ld-c-50/

Says conductivity is 0.039 W·m⁻¹·K⁻¹